Long Day's Journey into <Bleep>

[C. P. Ghost]

On Thu, Jun 9, 2011 at 5:53 AM, Chad Perrin <perrin at apotheon.com> wrote:
> On Wed, Jun 08, 2011 at 05:56:59PM -0700, Gary Kline wrote:
>>
>> I'm still bringing back the dozens of things I removed from ethic.
>> And testing new ideas.  But I have a general question: have any of
>> you wizards who run your own domains or otherwise use a switch [or
>> hub] *ever* had it just-quit?!  It is solid-state.  Yes, the box is
>> within my feet/foot reach.  I have accidently kicked it i suppose,
>> but still.
>
> I think I've just had ports die one by one on a switch until it no longer
> worked.  I don't think I've ever had the whole thing go poof for no
> evident reason.

Same here... a lot of times.

My last experience with a dying port on a switch was a few days ago
while JumpStart-ing Solaris via OBP. The process hung everywhere
from RARP, BOOTP, TFTP and NFS... until we figured out the port
on the switch was slowly dying.

Funny thing was that this problem was masked by TCP's error correction
mechanisms for quite some time and became only critical with UDP: the
TCP connections were slow as hell, but since the machine wasn't used for
high throughput anyway, the local junior admin assumed it was some kind
of software/hardware error on the host. She saw the many input errors (Ierrs)
in netstat -i, but didn't know what to do about them. ;-)

So yes, switches rarely stop altogether, the ports usually degrade, one
by one.

> Chad Perrin [ original content licensed OWL: http://owl.apotheon.org ]

-cpghost.

-- 
Cordula's Web. http://www.cordula.ws/